Scapular Dyskinesis, Glenohumeral Internal Rotation, and Pectoralis Minor Length in Male Elite Tennis Players

INTRODUCTION

The scapulothoracic articulation contributes to glenohumeral stability and function during repetitive overhead actions such as the tennis serve and groundstrokes.¹ During the serve, the shoulder moves into approximately 90° abduction and maximal external rotation (ER), followed by rapid internal rotation (IR) and flexion, placing mechanical stress on the subacromial space and causing cumulative microtrauma.² Thus, functional stability relies on dynamic scapular control, and alterations in its motion or positioning are major risk factors for shoulder injury.^3–5

Shoulder complex injuries constitute 25–47.7% of all upper-extremity injuries in tennis players.⁶ Altered activation of scapular stabilizers produces abnormal motion patterns known as scapular dyskinesis (SD).^7,8 Electromyographic evidence from individuals with SD shows upper trapezius overactivity, delayed lower trapezius recruitment, and decreased serratus anterior activation.,^7,9,10 disrupting optimal scapular alignment and motion efficiency.^11,12 A meta-analysis reported that asymptomatic athletes with SD had an increased risk of developing shoulder pain (RR = 1.43, 95% CI = 1.05–1.93),¹³ although its causal contribution remains uncertain because SD may be predisposing or compensatory.⁸ Glenohumeral internal-rotation deficit (GIRD) is characterized by reduced IR and has been associated with increased scapular anterior tilt, a kinematic pattern commonly observed in SD.^14,15

GIRD and pectoralis minor shortening have been described as structural factors related to SD.^12,16,17 In overhead athletes, GIRD has been characterized by substantial IR loss and, in more severe cases, a concomitant reduction in total rotational motion.¹⁸ The dominant shoulder in overhead athletes has been reported to exhibit posterior capsular hypertrophy associated with GIRD,^19,20 promoting a scapular “wind-up” posture that predisposes overhead athletes to SD.^12,21 A shortened pectoralis minor increases anterior tilt and downward scapula rotation, narrowing the subacromial space.^12,22,23 Yesilyaprak et al. reported that each one-unit decrease in the PMI increased the odds of SD, assessed via clinical observation, by 96%, indicating a strong relationship between anterior soft-tissue tightness and scapular malalignment.¹⁶ These adaptations resemble GIRD-related scapular wind-up, suggesting that both mechanisms may interact to produce SD via complementary structural pathways.^16,21,23

Repetitive overhead actions can alter scapular alignment in tennis players.^24,25 Among elite players, SD prevalence has been reported in 57.7% of dominant and 45.9% of non-dominant shoulders.²⁶ Although scapular malalignment is a recognized potential precursor to shoulder pathology, few studies have examined how SD grade relates to glenohumeral rotational motion and pectoralis minor morphology. Previous investigations have primarily examined SD prevalence²⁷ or isolated factors such as GIRD or the PMI.^28,29 Therefore, the aim of this study was to identify the prevalence and distribution of SD grades in elite tennis players and to investigate their association with glenohumeral rotational ROM and PML. The authors hypothesized that the dominant shoulder would exhibit a higher SD prevalence and that higher SD grades would correspond to reduced IR and TROM and shorter PML.

MATERIALS AND METHODS

PARTICIPANTS

This cross-sectional study included 50 male elite tennis players (between the ages of 14–24 years) registered with the Korea Tennis Association, were recruited through voluntary response to study advertisements posted at training facilities. Seven players met the predefined exclusion criteria due to having shoulder injuries limiting current training or competition participation. Therefore, 43 participants were included in the final analyses. All participants were national-level competitors who regularly competed and voluntarily provided written informed consent. For minors, additional consent was obtained from a parent or legal guardian. Eligibility criteria required active participation in structured training sessions at least four times per week for ≥1.5 h per session.²⁹ This training threshold was based on prior reports indicating an average weekly training volume of approximately 7–10 h/week for youth elite tennis players and was applied to both adolescent and adult athletes. Players with shoulder or upper-extremity injury or surgery within the previous three to six months that limited current competition participation were excluded. Sample size estimation using G*Power software (version 3.1; University of Kiel, Germany) was based on differences in glenohumeral IR across SD grades, using parameters derived from prior studies (effect size = 0.5,³⁰ α = 0.05, power = 0.80), yielding a minimum of 42 participants. The study was approved by the Institutional Review Board of (1044308-202502-HR-233-02) and conducted in accordance with the Declaration of Helsinki.

PROCEDURES

Participant demographic data (age, height, body mass, injury history, and training volume) were collected via questionnaire. The dominant side was defined as the serving arm. Each player completed the scapular dyskinesis test (SDT) and lateral scapular slide test (LSST) to evaluate dynamic and static scapular motion, respectively, followed by assessments of passive shoulder ROM (IR, ER, and TROM) and the PMI. The SDT was administered by a single examiner, and the results were concealed from the examiner conducting the LSST, ROM, and PMI measurements to minimize bias and maintain a single-blind design. Intra-rater reliability for all measurements was assessed in a separate session, independent of the main data collection, using repeated measurements obtained with a one-week interval. SDT grades (normal, subtle, obvious) were used as the independent variable, and all physical measures were analyzed separately for each shoulder.

Scapular Dyskinesis Test (SDT)

The SDT was used to evaluate dynamic scapular motion. Participants held dumbbells scaled to body weight (1.4 kg if <68.1 kg; 2.3 kg if ≥68.1 kg) and elevated both arms from 0° to 180° over a standardized 3 second cadence, paced by a metronome, for five repetitions while standing with thumbs pointing upward.³¹ The examiner observed scapular motion from 2 meters away and classified it as normal, subtle, or obvious according to standardized criteria.^31,32 The SDT shows excellent inter-rater reliability (κ = 0.86),³² and intra-rater reliability in this study was κ = 0.72–0.82.

Lateral Scapular Slide Test (LSST)

The LSST was conducted using a precision vernier caliper (CAS, Yangju, Republic of Korea) to measure the distance between the T7 spinous process and the scapular inferior angle in three positions: (1) arms relaxed, (2) hands on pelvis with thumbs pointing posteriorly, and (3) 90° abduction with full IR.¹ Each position was measured twice, and the mean value was used for analysis.³³ The examiner identified the C7 spinous process and subsequently measured the distance from T7 to the scapular inferior angle.³³ Previous studies have shown excellent LSST reliability (ICC >0.92),³⁴ and intra-rater reliability in this study was ICC = 0.87–0.94.

Shoulder Range of Motion (ROM)

Passive shoulder IR and ER were measured bilaterally in the supine position using a digital inclinometer (EX-POWER, Ansan, Republic of Korea) placed at the midpoint of the distal forearm.³⁵ One examiner stabilized the coracoid process and the spine of the scapula,³⁶ while another passively rotated the humerus to the end range defined as the first onset of scapular movement, without the application of overpressure.³⁷ Each measurement was performed twice, and the mean value was used for analysis.³⁵ TROM was calculated as the sum of IR and ER.³⁸ This procedure has demonstrated excellent reliability in previous studies (ICC = 0.96–0.99),³⁷ and intra-rater reliability in this study was ICC = 0.89–0.93.

Pectoralis Minor Length (PML)

PML was measured in the supine position.²⁸ The linear distance between the coracoid process and the fourth rib at the sternocostal junction was measured using a precision vernier caliper (CAS, Yangju, Republic of Korea).³⁹ Measurements were performed with the arms resting at the side and the palms facing upward. Each measurement was performed twice, and the mean value was used for analysis.¹⁶ To standardize for body size, the measured distance was divided by the participant’s height and multiplied by 100 to calculate the normalized PMI.²³ Previous studies have reported good reliability for this procedure (ICC = 0.82–0.87),³⁹ and intra-rater reliability in this study was ICC = 0.89.

STATISTICAL ANALYSES

All analyses were performed using IBM SPSS Statistics version 29.0 (IBM Corp., Chicago, IL, USA). Differences in SDT grade distribution between dominant and non-dominant shoulders were examined using Bowker’s test of symmetry for paired categorical data. Separate one-way analyses of variance (ANOVA) were conducted for each dependent variable (IR, ER, TROM, and PMI) to compare outcomes across SDT grades (normal, subtle, and obvious), followed by Bonferroni post hoc comparisons. Paired-samples t-tests were used to compare dominant and non-dominant shoulders for IR, ER, TROM, and PMI, and LSST values were analyzed using one-way ANOVA. Statistical significance was set at p < 0.05. Effect sizes (η² for ANOVA and Cohen’s d for t-tests) were interpreted using established thresholds. For η², values of 0.01, 0.06, and 0.14 were considered small, medium, and large effects, respectively. For Cohen’s d, values of 0.2, 0.5, and 0.8 were used to define small, medium, and large effects, respectively.

RESULTS

Forty-three elite tennis players were included in the analysis. No significant differences were found in body mass (p = 0.524, d = 0.31) or height (p = 0.422, d = 0.39) between the normal and SD groups. Across 86 shoulders, 70.9% (n = 61) demonstrated SD. In the dominant shoulder, normal scapular motion occurred in 20.9% (n = 9), subtle SD in 53.5% (n = 23) and obvious SD in 25.6% (n = 11). In the non-dominant shoulder, normal scapular motion was observed in 37.2% (n = 16), subtle SD in 51.2% (n = 22), and obvious SD in 11.6% (n = 5). Bowker’s test of symmetry demonstrated a significant asymmetry in SD grade distribution between dominant and non-dominant shoulders (χ² = 38.701, p < 0.001) (Table 1).

In the dominant shoulder, IR ROM differed significantly across SDT grades (F = 13.548, p <0.001, η² = 0.40). Post hoc testing revealed greater IR ROM in the normal group than in the subtle (p = 0.011) and obvious (p <0.001) groups, and greater IR ROM in the subtle group than in the obvious group (p = 0.012). ER ROM did not differ among groups (p = 0.569). TROM differed significantly across SDT grades (F = 4.858, p = 0.013, η² = 0.20), with the normal group showing greater TROM than the obvious group (p = 0.014) (Table 2).

In the non-dominant shoulder, IR ROM significantly decreased with increasing SDT grade (F = 18.233, p < 0.001, η² = 0.48). TROM also differed across grades (F = 6.572, p = 0.003, η² = 0.20), with the normal group showing greater TROM than both subtle (p = 0.039) and obvious (p = 0.005) groups. ER ROM showed no significant differences (p = 0.392) (Table 3).

When comparing sides, the dominant shoulder exhibited significantly lower IR ROM (p < 0.001, d = 0.62) and greater ER ROM (p = 0.004, d = 0.47), with no significant difference in TROM (p = 0.816, d = 0.04) (Table 4).

The PMI decreased significantly as SDT grade increased. PMI differed significantly across SDT grades in both the dominant (F = 19.625, p <0.001, η² = 0.50) and non-dominant shoulders (F = 19.432, p <0.001, η² = 0.49). Post hoc analyses revealed lower PMI values in the obvious group than in the normal and subtle groups (p <0.05), indicating greater pectoralis minor tightness (Table 5).

The dominant shoulder demonstrated significantly lower PMI than the non-dominant shoulder (p < 0.001, d = 0.56) (Table 6).

In the dominant shoulder, LSST distances did not differ significantly among SDT grades across all three testing positions. At Position 1, mean distances were 9.69 ± 1.00 cm (normal), 9.19 ± 1.26 cm (subtle), and 8.69 ± 0.95 cm (obvious) (F = 1.911, p = 0.161, η² = 0.09). Position 2 yielded values of 9.98 ± 0.98 cm, 9.61 ± 1.46 cm, and 8.98 ± 1.10 cm (F = 1.570, p = 0.221, η² = 0.07). Position 3 produced 10.05 ± 0.93 cm, 9.89 ± 1.17 cm, and 9.33 ± 1.11 cm (F = 1.236, p = 0.301, η² = 0.06), again showing no significant group differences (Table 7).

The non-dominant shoulder showed a similar trend, with no significant differences among groups. At Position 1, mean distances were 9.21 ± 0.93 cm, 8.88 ± 1.08 cm, and 8.89 ± 0.80 cm (F = 0.520, p = 0.599, η² = 0.03). At Position 2, means were 9.86 ± 1.06 cm, 9.51 ± 1.26 cm, and 9.39 ± 1.17 cm (F = 0.509, p = 0.605, η² = 0.02). At Position 3, means were 10.41 ± 1.27 cm, 9.66 ± 1.03 cm, and 9.64 ± 0.95 cm (F = 2.242, p = 0.119, η² = 0.10), with none reaching statistical significance (Table 8).

DISCUSSION

PREVALENCE AND SPORT-SPECIFIC ADAPTATIONS OF SD

SD occurred in 70.9% of all shoulders (dominant: 79.1%; non-dominant: 62.8%). The proportion of obvious SD in the dominant shoulder was more than twofold that of the non-dominant shoulder (25.6% vs. 11.6%). Previous studies have also found greater dominant-side prevalence in professional tennis players (57.7% vs. 45.9%),²⁶ although these investigations assessed only SD presence without grading. The higher prevalence observed in the current cohort may reflect differences in competitive level, training volume, age, or methodological factors, including the use of graded rather than binary assessment. Repetitive overhead loading induces posterior capsular tightness and asymmetric muscle activation, contributing to decreased scapular upward rotation and increased anterior tilt.^10,40 These kinematic alterations are biomechanical mechanisms underlying SD.^5,25 Comparable evidence from elite handball athletes showed that obvious SD, reduced IR ROM, and ER weakness were significantly associated with shoulder injury.³⁵ However, side-to-side differences may partly represent non-pathological performance adaptations, warranting cautious interpretation of their clinical significance.⁴¹ The observed asymmetry may be related to cumulative mechanical loading from repetitive racket swings. Future longitudinal studies are required to determine whether repetitive unilateral loading contributes to SD-related symptoms or injury risk.

RELATIONSHIP BETWEEN SD GRADE AND GLENOHUMERAL ROTATION

IR and TROM significantly decreased with increasing SD grade in both shoulders, with the greatest restriction observed in the obvious SD group (p = 0.014). GIRD has been described as a loss of internal rotation in overhead athletes, often defined as a side-to-side deficit of approximately 18–20°.^42,43 In the present study, internal rotation decreased with increasing SD grade in both shoulders. Side-to-side comparison demonstrated reduced IR and increased ER in the dominant shoulder, while total rotational motion did not differ between sides. This pattern of IR loss with preserved total rotational motion is consistent with rotational adaptations reported in overhead athletes.^44,45 Posterior capsular tightness, humeral retroversion, and posterosuperior humeral head translation can all contribute to this pattern.¹⁹ Thomas et al. reported that dominant-side GIRD in pitchers co-occurred with reduced scapular upward rotation and increased protraction.⁴⁰ López-Vidriero et al. observed GIRD in 87% and SD in 57% of professional tennis players, implicating repetitive overhead loading as a major contributor.²⁶ Similarly, the present study demonstrated greater IR restriction in the obvious SD group, indicating that increasing SD grade is associated with greater rotational imbalance. The large effect size observed for IR further supports the clinical relevance of this finding. Future longitudinal studies are needed to clarify the relationship between posterior capsular adaptations, GIRD, and SD grade.

COMBINED EFFECTS OF SD AND INTERNAL ROTATION ASYMMETRY ON SHOULDER FUNCTION

Previous studies have reported associations between SD, IR asymmetry, and shoulder function.^13,19 SD reflects abnormal scapular motion, particularly excessive anterior tilt and protraction, which narrows the subacromial space and increases rotator-cuff impingement risk.¹² Ellenbecker et al. reported that male elite tennis players demonstrated an 8.72-mm greater anterior shoulder position on the dominant side compared with the non-dominant side. The reduced PMI observed in the dominant shoulder in the present study suggests pectoralis minor shortening may be associated with this increased anterior positioning.⁴⁶ In tennis players, a greater SD grade was associated with IR reduction, likely reflecting posterior capsular tightness and malalignment induced by repetitive overhead loading.^14,47 Borich et al. found increased scapular anterior tilt at end-range IR in athletes with GIRD.¹⁴ López-Vidriero et al. observed decreased IR, increased ER, and altered TROM in dominant shoulders of adolescent tennis players, suggesting early adaptive asymmetry.⁴⁸ A recent systematic review by Ellenbecker et al. (2024) reported that male tennis players typically exhibit dominant-side internal rotation loss ranging from −15.3° to −3.0°.⁴⁹ This range provides a reference framework for interpreting the internal rotation deficits observed across SD grades in the present study. In the present study, an approximate 18° reduction in internal rotation was observed across SD grades, exceeding the reported range. Hickey et al. noted that asymptomatic athletes with SD had an elevated risk of future shoulder pain (relative risk = 1.43, 95% CI 1.05–1.93).⁵⁰ The observed reductions in IR across SD grades are consistent with previous reports describing altered scapular kinematics in athletes with SD.⁵¹ Owing to the cross-sectional design of the current study, causality cannot be inferred. However, these adaptations may represent early functional adaptations associated with an increased risk of shoulder injury. Clinically, these findings emphasize the importance of assessing SD and IR asymmetry during assessments and rehabilitation of overhead athletes.

PECTORALIS MINOR SHORTENING AND SCAPULAR KINEMATICS

PML decreased significantly with increasing SDT grade in both shoulders (p <0.001), with the greatest shortening in the obvious SD group. Similar patterns have been reported in elite boxers with SD⁵² and in asymptomatic adults with PML ≤6.56 cm, who demonstrated a 96% positive SDT rate.¹⁶ The pectoralis minor promotes scapular anterior tilt and downward rotation, and its shortening may restrict posterior tilt and upward rotation during overhead tasks, potentially reducing subacromial space.^23,53 An association between pectoralis minor tightness and SD has been identified by other researchers.^16,23,52 Umehara et al. demonstrated that releasing this stiffness increased posterior tilt and external rotation during arm elevation.⁵⁴ Although it is unclear whether PML shortening is causal or adaptive,⁵⁵ the present findings suggested a strong association between anterior soft-tissue tightness and scapular kinematics. Future research should determine whether improving PML can influence scapular motion and SD severity, and clinically, PML shortening should be recognized as a modifiable factor during evaluation and intervention.

INTERPRETATION OF LSST FINDINGS AND MEASUREMENT LIMITATIONS

LSST measurements did not significantly differ among SDT grades at any position. The LSST, a static alignment-based assessment tool, may not adequately capture the three-dimensional scapular kinematics observed during actual sport-specific movements.^33,34 LSST is insensitive to muscle-length changes because its measured distance can be influenced by isometric activation of the serratus anterior and trapezius.^56,57 Laudner et al. found no change in scapular upward rotation following an intervention to increase PML, suggesting that muscle-length alterations do not necessarily modify static scapular alignment.⁵⁸ Repetitive, high-velocity overhead movements can induce posterosuperior humeral translation and dynamic instability,⁵⁹ which static scapulothoracic distance measures cannot effectively assess.¹⁵ Therefore, the LSST may not adequately reflect scapular positional differences associated with pectoralis minor shortening, highlighting the importance of incorporating dynamic assessments in clinical evaluation.

LIMITATIONS

This study had several limitations. First, the sample consisted exclusively of male elite tennis players aged 14–24 years, restricting generalizability to athletes of other ages, sexes, and performance levels. Second, uneven SD subgroup sizes, particularly the disproportionately large subtle SD group, may have reduced statistical power, highlighting the need for more balanced sampling in future studies. Finally, injury history and pain data were not collected, influencing the interpretation of the results.

CONCLUSION

The results of this study demonstrate a high prevalence of SD among elite male tennis players. Although overall SD presence was high in both shoulders, the proportion of obvious SD was more than two-fold higher in the dominant shoulder (25.6%) than in the non-dominant shoulder (11.6%). Higher SD grades were associated with reduced internal rotation, decreased total rotational motion, and shorter pectoralis minor length. These findings are consistent with shoulder rotational adaptations previously described in overhead athletes. Although causality cannot be established owing to the cross-sectional design, these findings may represent early functional changes associated with increased shoulder injury risk. Longitudinal and interventional studies are needed to clarify whether modifying SD influences shoulder function and injury development.

Conflict of Interest

The authors report no conflicts of interest.